BAGI Video Gas Leak Visualization

FEW 5704

Goal
This project was initiated to create a more effective method of organic vapor leak detection at oil refineries, thus increasing the efficiency with which regulatory compliance can be met by industry.

Performers
Sandia National Laboratories (SNL)
Livermore, CA

DOE Office of Industrial Technology (OIT)
Washington, DC

American Petroleum Institute (API)
Washington, DC

Texas Council on Environmental Quality
Houston, TX

Laser Imaging Systems (LIS)
Punta Gorda, FL

Environmental Protection Agency (EPA)
Washington, DC

Shell Oil Company
Houston, TX

Results
In the course of this project, an operator-portable backscatter-absorption gas imaging (BAGI) system was developed to produce real-time video imagery of numerous organic vapors of interest to the petroleum (and other) industries. The imager was shown to operate effectively at several field trials at petroleum processing facilities in the United States and abroad. The results serve as the basis of an EPA rule-change application submitted by API requesting that gas imaging be allowed as an alternative work practice to the established EPA Method 21, which dictates how leak surveillance is conducted. That rule change is now under consideration.

Benefits
Hydrocarbons emitted by petroleum refineries generate smog and produce urban ozone. Thus the EPA actively regulates the emission of hydrocarbons by processing facilities. Leak detection and repair (LDAR) programs are operated by major facilities to comply with these regulations.

By making a gas plume "visible" to its operator, the gas imager offers an efficient and effective alternative to EPA Method 21, which describes the use of handheld "sniffer" sensors to detect leak presence and magnitude. Sniffers imprecisely infer leak flux (in grams per hour, or g/hr) from point concentration (parts per million, or ppm) measurements. Leaks are missed if the probed point does not intersect the vapor plume. Finally, LDAR programs using sensors that must be manually brought in close proximity to the leak point are laborious and expensive (~$1 million per year at large refineries). Gas imaging instantly covers a large area, thus ensuring that the leak point is in the measurement zone, allowing many points to be probed simultaneously, and providing a visual indication of leak flux.

Background
Hydrocarbons do not absorb light at visible wavelengths and are not seen by the eye. They do, however, absorb in the infrared (IR) and are viewable at specific wavelengths absorbed by the gas. Two methods of IR gas imaging exist-passive and active. The former views a scene using thermal IR radiation emitted by the gas and scene. The latter illuminates the scene using laser light absorbed by the gas as it images the scene using backscattered light. The intensity of the passive gas image depends on the temperature difference between the gas and scene, as well as on plume thickness and concentration. Active imaging has no temperature dependence and thus was selected to be used in this project.

Active gas imaging devices were being produced for commercial sale by LIS at the initiation of this project; however, those devices used CO2 lasers whose emission wavelengths (9-11 µm) are not absorbed by hydrocarbons relevant to the petroleum industry. Thus, a critical challenge faced by this project was the creation of a laser emitting at the mid-IR wavelengths (3.1-3.6 µm) that are absorbed by hydrocarbons, while meeting the size and efficiency requirements of an operator-portable system.

Summary
Project researchers have:

• Developed a van-mounted hydrocarbon gas imager.
• Demonstrated the imager's performance in a Texas oil refinery (April 1999).
• Developed an operator-portable, battery-powered hydrocarbon gas imager.
• Conducted laboratory tests in support of an EPA rule-change submission.
• Conducted the following field trials:
  - California petroleum refinery (January 2003).
  - Texas petroleum refinery (February 2003).
  - Petrochemical lab tests-BP at Naperville, IL (August 2003).
  - Southampton, U.K., petroleum refinery demonstration (September 2003).
  - Texas petroleum refinery test (January 2004).
  - Texas petrochemical plant tests (February 2004).

Infrared lasers for both the van-mounted and operator-portable imagers were constructed by coupling efficient nonlinear frequency converters to diode-pumped solid-state pump lasers. The former employed nonlinear optical crystals to convert light at the pump frequency to tunable radiation in the mid-IR. The first pump laser used was a water-cooled Nd:YAG laser suitable for use in the van-mounted configuration. The operator-portable imager employed a significantly smaller pump consisting of a miniature Nd:YAG laser amplified by a fiber-optic amplifier. This allowed creation of an operator-portable system capable of operating for >2 hours on battery power. At field trials, it was demonstrated to successfully image gas leaks at ranges up to 30 feet and detection limits corresponding to leak rates of ~2-60 g/hr of hydrocarbons.

Current Status (October 2005)
This project was completed in September 2004. At present, API and EPA are moving forward (using data generated in this project) with the rule change procedure to allow gas imaging as an alternative work practice. As a result of the success of this effort, several passive-camera manufacturers also have become involved in the testing. Presently, industry is favoring the passive cameras because of more immediate availability and lower cost and despite the thermal dependence of the gas image.

Funding
This project was funded through DOE's Natural Gas and Oil Technology Partnership Program.

Publications
Goers, U.B., Kulp, T.J., and McRae, T.G., Backscatter Absorption Gas Imaging of VOC's and its use in Petrochemical Leak Detection, published in the Proceedings of the 2000 Air and Waste Management Society Meeting, June 2000.

Siegell, Taback, McRae, and Kulp, Development of Smart LDAR for Fugitive Emissions Control, presentation at the Valve World 2000 Conference, June 2000.

Project Start: October 1, 1994
Project End: April 15, 2005

Anticipated DOE Contribution: $2,007,000
Performer Contributions: $325,000 (14% of total)

Other Government Organizations Involved
OIT, EPA

Contact Information
NETL - David Alleman (david.alleman@netl.doe.gov or 918-699-2057)
SNL - Thomas Kulp (thomas.kulp@snl.gov or 925-294-3676)

Diagram shows the function of the operator-portable imager. The IR laser beam is raster-scanned across the target to illuminate it. As the illumination occurs, a real-time video image of the scene is created using the backscattered laser light. If a gas is present in the scene that can absorb the laser light, it appears in the image as a dark cloud.

The operator-portable imager in use. The inset shows the gas image (propane, in this case) visible to the operator.

The shoulder-mounted BAGI imager in use at a Texas refinery.